deberta-v3-base-clickbait-task1-20-epoch-post

This model is a fine-tuned version of microsoft/deberta-v3-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.4780
• Accuracy: 0.7

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	200	0.7377	0.705
No log	2.0	400	0.7807	0.67
0.7497	3.0	600	0.7980	0.7075
0.7497	4.0	800	0.9537	0.7125
0.3256	5.0	1000	1.1896	0.6975
0.3256	6.0	1200	1.2719	0.7125
0.3256	7.0	1400	1.6965	0.7
0.1192	8.0	1600	1.7920	0.7075
0.1192	9.0	1800	2.0342	0.695
0.0497	10.0	2000	2.1345	0.71
0.0497	11.0	2200	2.1191	0.7075
0.0497	12.0	2400	2.3217	0.705
0.0288	13.0	2600	2.3254	0.7075
0.0288	14.0	2800	2.3810	0.7
0.0147	15.0	3000	2.3537	0.71
0.0147	16.0	3200	2.3389	0.6975
0.0147	17.0	3400	2.4556	0.69
0.01	18.0	3600	2.4238	0.705
0.01	19.0	3800	2.4632	0.695
0.0066	20.0	4000	2.4780	0.7

Framework versions

• Transformers 4.44.0.dev0
• Pytorch 2.4.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1